SSC (Shadowslice Snow Columns) 3x3x3 Method

[shadowslice e]

Guys, I think I have an idea to make this more efficient and better for the last few steps. Instead of doing belt + CO then CTL, then CPETL, then EPBL, which ends up being about 21 + 5 + 12 + 11 = 49 or so, why don't we do this:

1. Belt + CO (20 moves)
2. Solve FB in the bottom-left (3 or 4 moves)
3. CP + FD + BD (5 or 6 moves)
4. L11P, which would have 45 cases (~11 moves)

Total: 39 STM

What do you mean by CP? Like actual CP? Because that it think would be hard to recognise and probably would need algs to do efficiently.

I would be good with doing CTL+FD+BD but the recognition foe the last step could be messy (though definately doable) so a system may need to be created or some way to semi-predict the case.

 

[Shiv3r]

so who wants to finish up the alg set that crafto22 posted?

 

[adimare]

None of the example solves from the OP work.

 

[sqAree]

None of the example solves from the OP work.

scramble: B2 L D2 R' B2 U2 B2 F2 L B2 R' B2 R' B2 D' F2 U R2 U L' R2 F2
Solution: L B2 L U D' R' (orient edges + 3 E layer edges + orient 3 D corners)
D2 R U2 R' (set up WV)
U' L' U R U' R' L (WV)
U' R2 U2 R2 (sort corners)
U' D2 R2 U' R2 U D R2 D' R2 (permute corners)
U (UL/UR)
M2 D M' D2 M' D' (DL/DR)
D2 M D2 M' (LL)

At least this one works for me. Haven't checked the others.

 

[adimare]

You're right, one works, must have deleted something by accident when copy/pasting everything to alg.cubing.net.

Here are the other 2:
x2 //(in inspection)
(B' R D' R' F L D') //- EoLine/Edge
z //(because I'm not used to making an EoEdge yet)
R //(3rd edge)
D R2 //(3d corners)
M2 U2 M2 //(setting up WV)
U2 L' U2 R U' R' U2 L //(Fingertrick friendly WV http://www.cubewhiz.com/wv.php)
U R2 (D2 U') R2 //(corners to edges)
D' R2 (U D') R2 D R2 U' R2 U R2 U' R2 //(corner permutation- worst case http://www.cubezone.be/square1step4.html )
U M U2 M' D' //(LD and RD)
M2 U2 M' U2 M' //(UL and UR)
U M' U2 M'

z2 // Inspection
F' R' D' // EoEdge
z D2 R // 3rd e-slice piece
U R2 D // 3 corners
M2 U M2 // set up to SLS (Shadowslice Last Slot- new last slot- details below)
U2 R U R // SLS
U' R2 D R2 D' R2 U2 R2 // CTL (Corners To Layers) -absolutely awful worst case.
D' R2 (U D') R2 D R2 U' R2 U R2 U' R2 // CPBL (Corner Permutaion of Both Layers)- worst case.
D' M2 D // LD/LR permutation
M2 U' M' U2 M' U' M2 U // LU/RU permutation- worst case
M' U2 M2 U2 M //LFE (Last Four Edges)

Neither works.

 

[shadowslice e]

You're right, one works, must have deleted something by accident when copy/pasting everything to alg.cubing.net.

Here are the other 2:
x2 //(in inspection)
(B' R D' R' F L D') //- EoLine/Edge
z //(because I'm not used to making an EoEdge yet)
R //(3rd edge)
D R2 //(3d corners)
M2 U2 M2 //(setting up WV)
U2 L' U2 R U' R' U2 L //(Fingertrick friendly WV http://www.cubewhiz.com/wv.php)
U R2 (D2 U') R2 //(corners to edges)
D' R2 (U D') R2 D R2 U' R2 U R2 U' R2 //(corner permutation- worst case http://www.cubezone.be/square1step4.html )
U M U2 M' D' //(LD and RD)
M2 U2 M' U2 M' //(UL and UR)
U M' U2 M'

z2 // Inspection
F' R' D' // EoEdge
z D2 R // 3rd e-slice piece
U R2 D // 3 corners
M2 U M2 // set up to SLS (Shadowslice Last Slot- new last slot- details below)
U2 R U R // SLS
U' R2 D R2 D' R2 U2 R2 // CTL (Corners To Layers) -absolutely awful worst case.
D' R2 (U D') R2 D R2 U' R2 U R2 U' R2 // CPBL (Corner Permutaion of Both Layers)- worst case.
D' M2 D // LD/LR permutation
M2 U' M' U2 M' U' M2 U // LU/RU permutation- worst case
M' U2 M2 U2 M //LFE (Last Four Edges)

Neither works.

Do they work now? (sorry I can't check because my phone won't do alg.cubing.net)

 

[adimare]

Almost, in the second one I replaced the M's with Ms in the very last line and got it to work. here are alg.cubing.net links for all three:

Solve 1

Solve 2

Solve 3

 

[sqAree]

I plan to add some explained example solves to the wiki this weekend btw. Probably using the new CO method.

 

[TyeDye]

Almost, in the second one I replaced the M's with Ms in the very last line and got it to work. here are alg.cubing.net links for all three:

Solve 1

Solve 2

Solve 3

Thank you for getting all of the solves to work! I couldn't figure out where it would go wrong when I would do it.

 

[crafto22]

Okay I have a new idea for a better solution after Belt + CO:

1. Solve Roux blocks whilst preserving CO, which is about 10 or 12 moves on average
2. CPETL, so permute the corners and get the D-layer edges to D, which is about 11 moves average I think
3. Solve the cube with a 7-9 move alg

So total moves is about 43 STM

 

[Teoidus]

Are you sure solving D edges is most efficient? What about, say, solving ULUR?

 

[crafto22]

Are you sure solving D edges is most efficient? What about, say, solving ULUR?

Good idea, that sounds way better.

 

[Teoidus]

Hm. Or maybe it'd be even better to solve, say, a D edge and its opposite (so either DF and UB or DB and UF)?
That would leave pretty decent EO cases (either an arrow or a twoflip, which is always 4 moves away from arrow).
I'm just looking at this in terms of Roux LSE though, so it might not actually be themost efficient way.

 

[sqAree]

As this is SSC we assume full EO being already done anyway, right?

 

[crafto22]

As this is SSC we assume full EO being already done anyway, right?

Yes, EO is already done. One would simply use PLL algs to get an easy L4E case, so the whole step could be one-looked.
So these are the steps:
1. Belt + CO
2. Roux blocks
3a. PLL alg to solve UL and UR and influence other edges
3b. M' U2 M, M U2 M' or M2 U2 M2 to solve the cube

 

[wir3sandfir3s]

Yes, EO is already done. One would simply use PLL algs to get an easy L4E case, so the whole step could be one-looked.
So these are the steps:
1. Belt + CO
2. Roux blocks
3a. PLL alg to solve UL and UR and influence other edges
3b. M' U2 M, M U2 M' or M2 U2 M2 to solve the cube

Wouldn't this be easier than roux blocks, and really your whole idea in general?
1. 3/4 belt
2. Solve 3 D layer corners, then solve last corner and edge and CO with WV
3. Solve D Layer edges, insert last one while solving EO
4. PLL
1+2 can be combined to solving pairs or pseudo pairs

 

[crafto22]

Wouldn't this be easier than roux blocks, and really your whole idea in general?
1. 3/4 belt
2. Solve 3 D layer corners, then solve last corner and edge and CO with WV
3. Solve D Layer edges, insert last one while solving EO
4. PLL
1+2 can be combined to solving pairs or pseudo pairs

That isn't as efficient. Also, your idea doesn't work, you forgot to include separating and permuting the corners.

 

[sqAree]

Yes, EO is already done. One would simply use PLL algs to get an easy L4E case, so the whole step could be one-looked.
So these are the steps:
1. Belt + CO
2. Roux blocks
3a. PLL alg to solve UL and UR and influence other edges
3b. M' U2 M, M U2 M' or M2 U2 M2 to solve the cube

Example Solve

PLL is definitely not sufficient for this.

 

[sqAree]

That isn't as efficient. Also, your idea doesn't work, you forgot to include separating and permuting the corners.

He wants to solve 3 D-layer corners directly even before CO, so after inserting it with WV the corners are already separated and PLL at the end permutes them.

 

[wir3sandfir3s]

That isn't as efficient. Also, your idea doesn't work, you forgot to include separating and permuting the corners.

you don't need to separate them and permute them, have you tried it?
Also changing it for efficiency, really just the same thing:
1. Solve 3 D CE pairs.
2. Inset last pair with a much less move and alg intensive (maybe even intuitive) WV
3. Solve remaining D edges, last one should be inserted while doing EO
4. PLL
It works fine, and is decently efficient and DEFINITELY can be made into something more.
Some things to possibly change:
1. Permute corners after WV, so you are left with EPLL
2. After WV, solve Roux Blocks and then D layer edges? Basically the same though.
3. Influence F2L each time you insert an edge.
4. Maybe add some cool Algs to do some cool things?